Field experiments, in which the researcher manipulates one or more variables in a naturally occurring setting, have sometimes been used in studies of social conflict and should probably be used more often. They are more useful than observational studies for assessing the impact of novel conditions, establishing cause and effect, and reducing confounding. And they are more useful than laboratory experiments for examining long-term effects and those that involve strong passions, and for establishing external validity. However field experiments also have their limitations. Some variables cannot be practically or ethically manipulated and require the use of observational methods, which are also more useful for looking at the relationships among a large number of variables and for estimating the strength of association between variables. Furthermore, laboratory experiments allow more control of conditions and greater flexibility in designing manipulations. What this suggests is that all three methods have their value.